[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAD=FV=VZQVfb7=+xN-kVPk6yWeYmrfuMtd=fUjXp6e_JHByAGg@mail.gmail.com>
Date: Tue, 6 Sep 2022 15:28:41 -0700
From: Doug Anderson <dianders@...omium.org>
To: Yunlong Jia <yunlong.jia@....corp-partner.google.com>
Cc: LKML <linux-kernel@...r.kernel.org>,
Bob Moragues <moragues@...omium.org>,
Henry Sun <henrysun@...gle.com>,
Dmitry Torokhov <dmitry.torokhov@...il.com>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Guenter Roeck <groeck@...omium.org>,
Johnny Chuang <johnny.chuang.emc@...il.com>,
"open list:HID CORE LAYER" <linux-input@...r.kernel.org>
Subject: Re: [PATCH 2/2] input: touchscreen: elants_i2c: Add eth3915n
touchscreen chip
Hi,
On Fri, Sep 2, 2022 at 1:33 AM Yunlong Jia
<yunlong.jia@....corp-partner.google.com> wrote:
>
> The eth3915n requires more delay time than the eth3500 when reset.
> Define EKTH3915_RESET_DELAY_MSEC as the reset delay time of eth3915n,
> about 300ms.
>
> Signed-off-by: Yunlong Jia <yunlong.jia@....corp-partner.google.com>
>
> ---
>
> drivers/input/touchscreen/elants_i2c.c | 17 ++++++++++++++++-
> 1 file changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/input/touchscreen/elants_i2c.c b/drivers/input/touchscreen/elants_i2c.c
> index c9dd703b0c7d8..60ddb58281bdf 100644
> --- a/drivers/input/touchscreen/elants_i2c.c
> +++ b/drivers/input/touchscreen/elants_i2c.c
> @@ -116,6 +116,7 @@
>
> #define ELAN_POWERON_DELAY_USEC 500
> #define ELAN_RESET_DELAY_MSEC 20
> +#define EKTH3915_RESET_DELAY_MSEC 300
>
> /* FW boot code version */
> #define BC_VER_H_BYTE_FOR_EKTH3900x1_I2C 0x72
> @@ -133,6 +134,7 @@
> enum elants_chip_id {
> EKTH3500,
> EKTF3624,
> + EKTH3915,
> };
>
> enum elants_state {
> @@ -664,6 +666,7 @@ static int elants_i2c_initialize(struct elants_data *ts)
>
> switch (ts->chip_id) {
> case EKTH3500:
> + case EKTH3915:
> if (!error)
> error = elants_i2c_query_ts_info_ekth(ts);
> break;
> @@ -1361,7 +1364,17 @@ static int elants_i2c_power_on(struct elants_data *ts)
> if (error)
> return error;
>
> - msleep(ELAN_RESET_DELAY_MSEC);
> + if (ts->chip_id == EKTH3915)
> + /*
> + * There need delay 300ms for power on sequence.
> + * T1 + T2 + T3 >= 305 ms
> + * T1: 0<time<500us
> + * T2: >5ms
> + * T3: >300ms
> + */
> + msleep(EKTH3915_RESET_DELAY_MSEC);
> + else
> + msleep(ELAN_RESET_DELAY_MSEC);
>
> return 0;
> }
> @@ -1686,6 +1699,7 @@ static const struct i2c_device_id elants_i2c_id[] = {
> { DEVICE_NAME, EKTH3500 },
> { "ekth3500", EKTH3500 },
> { "ektf3624", EKTF3624 },
> + { "ekth3915", EKTH3915 },
> { }
> };
> MODULE_DEVICE_TABLE(i2c, elants_i2c_id);
> @@ -1702,6 +1716,7 @@ MODULE_DEVICE_TABLE(acpi, elants_acpi_id);
> static const struct of_device_id elants_of_match[] = {
> { .compatible = "elan,ekth3500", .data = (void *)EKTH3500 },
> { .compatible = "elan,ektf3624", .data = (void *)EKTF3624 },
> + { .compatible = "elan,ekth3915", .data = (void *)EKTH3915 },
Adding giant delays isn't my favorite, but it seems like the trend in
touchscreens is that newer ones take forever to power up. I guess this
is no exception. FWIW:
Reviewed-by: Douglas Anderson <dianders@...omium.org>
Powered by blists - more mailing lists